Fatigue life prediction of composite materials using polynomial classifiers and recurrent neural networks

Due to their massively parallel structure and ability to learn by example, artificial neural networks can deal with nonlinear problems for which an accurate analytical solution is difficult to obtain. These networks have been used in modeling the mechanical behavior of fiber-reinforced composite materials. Although promising results were obtained using such networks, more investigation on the appropriate choice of their structure and their performance in the presence of limited and noisy data is needed. On the other hand, polynomials networks have been known to have excellent properties as classifiers and are universal approximators to the optimal Bayes classifier. Not being dependant on various user defined parameters, having less computational requirements makes their use over other methods, such as neural networks, an advantage.

In this work, the fatigue behavior of unidirectional glass fiber/epoxy composite laminae under tension–tension and tension–compression loading is predicted using feedforward and recurrent neural networks. These predictions are compared to those obtained using polynomial classifiers. Experimental data obtained for fiber orientation angles of 0°, 19°, 45°, 71° and 90° under stress ratios of 0.5, 0 and –1 is used.

It is shown that, even when a small number of experimental data points is used to train both polynomial classifiers and neural networks, the predictions obtained are comparable to other current fatigue life-prediction methods. Also, polynomial classifiers are shown to provide accurate modeling between the input parameters (maximum stress, R-ratio, fiber orientation angle) and the number of cycles to failure when compared to neural networks.     

Power-draw prediction by random forest based on operating parameters for an industrial ball mill

Estimation of mill power-draw can play a critical role in economics, operation and control standpoints of the entire mineral processing plants since the cost of milling is the single biggest expense within the process. Thus, several empirical power-draw prediction models have been generated based on a combination of laboratory, pilot and full-scale measurements of different milling conditions. However, they cannot be used in industrial plants, where in full-scale operations, only not few numbers of input parameters used in those models are measured. Moreover, empirical models do not assess the relationship between input features. This investigation is going to introduce random forest (RF), as a predictive model, beside of its associated variable importance measures system, as a sensible means for variable selection, to overcome drawbacks of empirical models. Although RF as a powerful modeling tool has been used in several problem solving systems, it has not comprehensively considered in the powder technology areas. In this investigation, an industrial ball mill database from Chadormalu iron ore processing plant were used to develop a RF model and explore relationships between power-draw and other monitored operating parameters. Modeling results indicated that RF can highly improve the prediction accuracy of power-draw as compared to the regression as a typical method (R²: 0.98 vs. 0.60, respectively) and rank operational milling parameters based on their importance.     

Predicting future technological convergence patterns based on machine learning using link prediction

Scientometrics - Technological convergence among different industries is an important source of innovation and economic growth. In this study, we propose a new framework for predicting patterns of...     

Learning dispatching rules using random forest in flexible job shop scheduling problems

In this paper, we address the flexible job-shop scheduling problem (FJSP) with release times for minimising the total weighted tardiness by learning dispatching rules from schedules. We propose a random-forest-based approach called Random Forest for Obtaining Rules for Scheduling (RANFORS) in order to extract dispatching rules from the best schedules. RANFORS consists of three phases: schedule generation, rule learning with data transformation, and rule improvement with discretisation. In the schedule generation phase, we present three solution approaches that are widely used to solve FJSPs. Based on the best schedules among them, the rule learning with data transformation phase converts them into training data with constructed attributes and generates a dispatching rule with inductive learning. Finally, the rule improvement with discretisation improves dispatching rules with a genetic algorithm by discretising continuous attributes and changing parameters for random forest with the aim of minimising the average total weighted tardiness. We conducted experiments to verify the performance of the proposed approach and the results showed that it outperforms the existing dispatching rules. Moreover, compared with the other decision-tree-based algorithms, the proposed algorithm is effective in terms of extracting scheduling insights from a set of rules.     

Predicting product development directions for new product planning using patent classification-based link prediction

Scientometrics - Predicting the possible development directions of a product is useful for planning innovative products. Therefore, a systematic approach based on link prediction is proposed in...     

Doc2vec-based link prediction approach using SAO structures: application to patent network

Scientometrics - As the amount of documents has exploded in the Internet era, many researchers have tried to understand the relationships between documents and predict the links between similar but...     

Glioma brain tumor detection and segmentation using weighting random forest classifier with optimized ant colony features

The uncontrolled growth of cells in brain regions leads to the tumor regions and these abnormal tumor regions are scanned by magnetic resonance imaging (MRI) technique as an image. This paper proposes random forest classifier based Glioma brain tumor detection and segmentation methodology using feature optimization technique. The texture features are derived from brain MRI image and these derived feature set are now optimized by ant colony optimization algorithm. These optimized set of features are trained and classified using random forest classification method. This classifier classifies the brain MRI image into Glioma or non-Glioma image based on the optimized set of features. Furthermore, energy-based segmentation method is applied on the classified Glioma image for segmenting the tumor regions. The proposed methodology for Glioma brain tumor stated in this paper achieves 97.7% of sensitivity, 96.5% of specificity, and 98.01% of accuracy.     

The essential link between research and technology

The behaviour of CdSe and its components was studied during their sublimation from vapour sources made of various materials (molybdenum, tungsten, corundum, graphite and tantalum) and covered with perforated lids of various geometries. The temperature dependence of the density of evaporated molecules within the source was studied as a function of the evaporation rate and the geometrical shape of the vapour source. The dependences found demonstrate the influence of the source material and the source shape on the evaporation properties of CdSe.     

A link between short-range and long-range properties of random sphere packings

We present a high precision particle-by-particle 3D reconstruction of granular systems composed of monodispersed spheres (sphere packings); the experimental approach is based on magnetic resonance imaging techniques. Our measurements revealed a strong correlation between the volume defined by the distance to the first nearest neighbor and the long-range average density. The main contribution to the amplitude decay of the correlation function can be described as exponential rather than power law up to a range equal to 7 sphere diameters. No evidence of geometrical structural changes as a function of the density was observed and neither regular crystallites nor any other statistically significant structures could be ascribed to a specific local arrangement. We concluded that granular compaction is the result of a process through which the system changes the average size of local structures without changing their local geometrical characteristics. These conclusions are supported by two-body correlation functions and Voronoi polyhedra space decomposition. The results provide a different perspective on the mechanisms underlying compaction with respect to previous works, and allow to discriminate between the different existing theoretical approaches.     

Development of computer‐aided approach for brain tumor detection using random forest classifier

The nonlinear development of cells in brain region forms the abnormal patterns in brain in the form of tumors. It is necessary to detect and diagnose the brain tumors in an automated manner using computer‐aided approaches at large population areas. The noises in brain magnetic resonance image is detected and reduced as preprocessing steps and then grey level co‐occurrence matrix are now extracted from the preprocessed brain image. In this article, random forest classifier‐based brain tumor detection and segmentation methodology is proposed to classify the brain image into normal or abnormal. The proposed brain tumor detection and segmentation system is analyzed in terms of sensitivity, specificity, false‐positive rate, false‐negative rate, likelihood ratio positive, and likelihood ratio negative.     

Combined weakest link and random defect model for describing strength variability in fibres

A mathematical model which describes the strength variability along the length of a fibre was developed. The model is a combination of the modified weakest link and random defect models. This combined model describes very well the strength variability data of aramid fibres.Nomenclature L Specimen length - F(s) Cumulative frequency distribution of link strengths - 1 — F(s) Survival function of a link - F _L(s) Cumulative frequency distribution of strengths of a specimen of length L - 1 — F _L(s) Survival function of a specimen of length L - s Strength variable - s ₀ Fibre defect-free strength for a random defect or combined model - s ₁, s ₂... Fibre strength at the point of a defect - s ₁, s ₂ ... Strength a fibre must have at the location of the defect to have a strength of s at the location of the defect - Length of a hypothetical link in a weakest link model - ₂, ₂ ... Defect frequencies (mean number per unit length) - v ₁, v ₂ ... Defect severities, 0 v 1 - (s) Defect frequency distribution function defined in terms of the strength at the defect - (v) Defect frequency distribution function defined in terms of the defect severity - , Defect frequency distribution parameters (Equation 14) - a, b Weibull distribution parameters (Equation 4) - P(m) Probability that m defects will occur in a given specimen length - m Number of defects occurring - ¯s Mean strength - CV Coefficient of variation of strength     

Mixed mode fracture initiation and trajectory prediction under random stresses

A method is developed for estimating (i) confidence intervals on the initial direction of crack extension and (ii) the probability of crack initiation in plane stress and plane strain problems. The method accounts for the uncertainty in applied stresses, fracture toughness, and crack geometry. It is based on classical theories of linear fracture mechanics for homogeneous isotropic materials, a computer code for deterministic fracture mechanics analysis (FRANC) and first and second order structural reliability algorithms (FORM/SORM). Several examples are presented to demonstrate the use and generality of the proposed method for probabilistic fracture mechanics analysis.

---

Résumé On développe une méthode pour estimer les intervalles de confiance sur la direction initiale de propagation d'une fissure et la probabilité d'amorçage d'une fissure dans des problèmes de tension plane et de déformation plane. La méthode tient compte des incertitudes dans les contraintes appliquées, dans la ténacité à la rupture et dans la géométrie de la fissure. Elle est basée sur les théories classiques de la mécanique de rupture linéaire pour les matériaux homogènes et isothropes, sur un code de calcul pour une analyse déterministe de la mécanique de la rupture (FRANC), et sur des algorithmes de fiabilité structurelle du premier et du deuxième ordre (FORM/SORM). Divers exemples sont présentés pour démontrer l'utilisation et la généralisation de la méthode proposée pour l'analyse probabilistique de la mécanique de la rupture.

     

Diagnosis of disc bulge and disc desiccation in lumbar MRI using concatenated shape and texture features with random forest classifier

Disc bulge and disc desiccation are the most common abnormalities occurring in the spine, which leads to severe low back pain. Despite computer-aided automatic abnormality diagnostic imaging systems are available still there is a need for betterment in diagnostic accuracy and in processing time. Image processing with combined imaging features like shape and texture has given better diagnostic ability when compared with processing with individual features. In the present study, the desiccated and bulged Intervertebral Discs (IVDs) are diagnosed automatically by combining shape features extracted using Histogram of Oriented Gradients (HOG) and texture feature extracted using novel Local Sub-Rhombus Binary Relation Pattern (LS-RBRP) techniques with Random Forest (RF) classifier. The performance analysis projects that the RF with HOG+LS-RBRP has an overall better accuracy of 94.7% when compared with HOG (87%) and LS-RBRP (90.2%) with RF classifier separately in categorizing the normal IVD, disc bulge and disc desiccation in the lumbar spine MRI.     

VMD和随机森林在反应堆金属撞击信号识别中的应用研究

反应堆中脱落部件或金属遗留件的存在会与压力边界产生碰撞而严重影响反应堆的安全,为有效解决由于特征信息不足和模型构建不合理而导致的反应堆中金属撞击信号识别精度低的问题,提出了一种基于变分模态分解(variational mode decomposition,VMD)和随机森林算法(random forest,RF)的反应...     

The link between Industry 4.0 and lean manufacturing: mapping current research and establishing a research agenda

In recent years, Industry 4.0 has emerged as one of the most discussed concepts and has gained significant popularity in both academia and the industrial sector. Both Industry 4.0 and lean manufacturing utilise decentralised control and aim to increase productivity and flexibility. However, there have been few studies investigating the link between these two domains. This article explores this novel area and maps the current literature. This is achieved through a systematic literature review methodology, investigating literature published up to and including August 2017. This article identifies four main research streams concerning the link between Industry 4.0 and lean manufacturing, and a research agenda for future studies is proposed.     

Quality loss prediction of mechanical system life considering linear degeneration and random error

To accurately predict quality loss of high reliability and long-life product in service, a dynamic quality characteristics model is established based on linear degeneration and random error. According to this model, the expectation and variance of dynamic quality characteristics can be figured out. Moreover, assuming that the quality characteristics follow the normal distribution at the initial stage, the expectation of quality loss and the life distribution of dynamic quality characteristics follow three types are derived, such as L-type (Larger is better), S-type (Smaller is better), and N-type (Nominal is better). According to the discount theory of quality loss, the present value model of dynamic quality loss based on the life is derived by combining life and the present value of dynamic quality loss. Then, this model is used to evaluate the quality loss of a GaAs laser and the rationality of the model is analyzed.     

Detection of pneumonia in chest X‐ray images by using 2D discrete wavelet feature extraction with random forest

Pneumonia is one of the most common and fatal diseases in the world. Early diagnosis and treatment are important factors in reducing mortality caused by the aforementioned disease. One of the most important and common techniques to diagnose pneumonia disease is the X‐ray images. By evaluating these images, various machine‐learning methods are used for accuracy in diagnosis. The presented study in this article utilizes machine‐learning techniques to evaluate these X‐ray images. The diagnosis of pediatric pneumonia is classified with a proposed machine learning method by using the chest X‐ray images. The proposed system firstly utilizes a two‐dimensional discrete wavelet transform to extract features from images. The features obtained from the wavelet method are labeled as normal and pneumonia and applied to the classifier for classification. Besides, Random Forest algorithm is used for the classification technique of 5856 X‐ray images. A 10‐fold cross‐validation method is used to evaluate the success of the proposed method and to ensure that the system avoided overfitting. By using various machine learning algorithms, simulation results reveal that the Random Forest method is proposed and it gives successful results. Results also show that, at the end of the training and validation process, the proposed method achieves higher success with an accuracy of 97.11%.     

Modelling non-Gaussian random fields using transformations and Hermite polynomials

In the context of modelling residual roughness on nominally flat moderately polished metal surfaces, a method is proposed for solving problems related to sample function properties and/or special points such as maxima, minima, saddle points for random fields having non-Gaussian height distributions by recasting them in terms of the corresponding problems for the much more tractable Gaussian random fields by means of transformations. Special reference is made to the expansion of the transformations in series of Hermite polynomials. While the use of Hermite polynomials in connection with transformations of random fields and the useful results they yield with regard to covariance functions are well known, this paper derives the most general explicit formula for the expectation of any product of several Hermite polynomials in correlated Gaussian arguments thereby allowing their application to the higher moments of the transformed random field, in particular, to the third moment, which may be used to measure skewness.     

Exploring precrash maneuvers using classification trees and random forests

Taking evasive actions vis-à-vis critical traffic situations impending to motor vehicle crashes endows drivers an opportunity to avoid the crash occurrence or at least diminish its severity. This study explores the drivers, vehicles, and environments’ characteristics associated with crash avoidance maneuvers (i.e., evasive actions or no evasive actions). Rear-end collisions, head-on collisions, and angle collisions are analyzed separately using decision trees and the significance of the variables on the binary response variable (evasive actions or no evasive actions) is determined. Moreover, the random forests method is employed to rank the importance of the drivers/vehicles/environments characteristics on crash avoidance maneuvers. According to the exploratory analyses’ results, drivers’ visibility obstruction, drivers’ physical impairment, drivers’ distraction are associated with crash avoidance maneuvers in all three types of accidents. Moreover, speed limit is associated with rear-end collisions’ avoidance maneuvers and vehicle type is correlated with head-on collisions and angle collisions’ avoidance maneuvers. It is recommended that future research investigates further the explored trends (e.g., physically impaired drivers, visibility obstruction) using driving simulators which may help in legislative initiatives and in-vehicle technology recommendations.